CCBUG: 391200
Details
- Reviewers
ngraham broulik pino elvisangelaccio cfeck - Group Reviewers
Dolphin - Commits
- R318:6e04f50081a9: Remove custom icon selection for trash
- Open Dolphin
- Edit trash in places item
Result: No custom icon option available
Diff Detail
- Repository
- R318 Dolphin
- Lint
Lint Skipped - Unit
Unit Tests Skipped
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
149 | Let's do this a little differently to conform to KDE style guidelines: if (m_url.scheme() == QLatin1String("trash")) { formLayout->addRow(i18nc("@label", "Choose an icon:"), m_iconButton); } else { m_iconButton->setVisible(false); } |
Also, can you add CCBUG: 391200 to the Summary section and edit your Test Plan to indicate that it should be done in Dolphin?
yeah sure, btw in the inline comment I think you meant .... if (m_url.scheme() != QLatin1String("trash")) instead of ==
Er, yeah, oops. Or just reverse the ordering of the statements inside the if and else blocks.
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
149 | Or even better, not create the KIconButton at all, in that case -- note you will need to check all the usages of m_iconButton. |
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
150 | Indent this line |
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
149 | +1 for not even creating it at all, if that's feasible. |
This works fine for me, and I think the code is sensible. It's quite a mess to not set m_iconButton at all in this case, so I'm okay with the patch as-is.
Let's wait for a final opinion from another Dolphin person.
It is a good start, but not enough:
- PlacesItemEditDialog::icon() will crash now, so its usage of m_iconButton must be guarded
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
147 ↗ | (On Diff #38449) | No need to SCREAM in comments :) |
This will crash as soon as you try to edit the Trash place (because m_iconButton in PlacesItemEditDialog::icon() won't be initialized).
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
67–71 ↗ | (On Diff #38449) | This is slightly overkill: return m_iconButton ? m_iconButton->icon() : QString(); is more than enough. |
No need to rush with the updates, please take your time to test your changes.
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
60 ↗ | (On Diff #38449) | This is unused, and it will even fail to build (since there is no more class declaration for it). |
Please, please, please:
- again, no rush with changes
- read again your own changes, because it looks like there is even no review by the own patch author
- TEST your changes
src/panels/places/placesitemeditdialog.cpp | ||
---|---|---|
62 ↗ | (On Diff #38449) | And actually this must return m_icon in case m_iconButton is null, to look like there was no icon change. |
Thanks, this now works, doesn't crash, and setting the icon still works for other locations. Code looks sane to me, but is this also good to you guys now, @pino, @broulik, and @elvisangelaccio?
This looks fine to me, and for consistency with KIO commit, we should ship 18.08 with this commit. Unless there are objections, or someone beats me to it, I will commit it to 18.08 branch in the coming days.
This does not make any sense to me. This looks already with the acks needed, doesn't it?
The current status is "Needs Review" up at the top of the page because you're still marked as a "Request Changes". Phabricator doesn't mark it as accepted as long as there's anyone who's still requesting changes.
Then fix Phabricator, since my "Request changes" is not for the latest version (which has only acks), but for previous ones.
I can't change Phabricator, but you can adapt to its quirks as we all have. It's just two clicks.
... that change the role I have in this review. Sorry, I will not do that. You have already all the data needed, including the fact that the patch is acked already. Unless you really care about the green tick, just to see that "everybody accepted it".
There are two more options:
- You "resign" as a reviewer from the revision.
- Someone else "removes" you as a reviewer.
Regardless, I'll land this on the stable branch now (per @cfeck) since there don't appear to be any remaining change requests.